[MRG] Reorganise under-sampling methods (#277)

* MAINT change organisation under-sampling

* FIX conflict api doc

* DOC add entry in whats new

* FIX indent in the documentation

* FIX add current module for linking the documentation

* FIX addres christos comments

* ENH Move the tests

* FIX remove useless import

Author: glemaitre

doc/api.rst

@@ -13,14 +13,34 @@ Under-sampling methods
     :no-members:
     :no-inherited-members:
 
-Classes
--------
+.. currentmodule:: imblearn
+
+Prototype generation
+--------------------
+
+.. automodule:: imblearn.under_sampling.prototype_generation
+   :no-members:
+   :no-inherited-members:
+
 .. currentmodule:: imblearn
 
 .. autosummary::
    :toctree: generated/
 
    under_sampling.ClusterCentroids
+
+Prototype selection
+-------------------
+
+.. automodule:: imblearn.under_sampling.prototype_selection
+   :no-members:
+   :no-inherited-members:
+
+.. currentmodule:: imblearn
+
+.. autosummary::
+   :toctree: generated/
+
    under_sampling.CondensedNearestNeighbour
    under_sampling.EditedNearestNeighbours
    under_sampling.RepeatedEditedNearestNeighbours
@@ -32,7 +52,6 @@ Classes
    under_sampling.RandomUnderSampler
    under_sampling.TomekLinks
 
-
 .. _over_sampling_ref:
 
 Over-sampling methods
@@ -42,8 +61,6 @@ Over-sampling methods
     :no-members:
     :no-inherited-members:
 
-Classes
--------
 .. currentmodule:: imblearn
 
 .. autosummary::
@@ -63,8 +80,6 @@ Combination of over- and under-sampling methods
     :no-members:
     :no-inherited-members:
 
-Classes
--------
 .. currentmodule:: imblearn
 
 .. autosummary::
@@ -83,8 +98,6 @@ Ensemble methods
     :no-members:
     :no-inherited-members:
 
-Classes
--------
 .. currentmodule:: imblearn
 
 .. autosummary::
@@ -105,18 +118,10 @@ Pipeline
 
 .. currentmodule:: imblearn
 
-Classes
--------
 .. autosummary::
    :toctree: generated/
 
    pipeline.Pipeline
-
-Functions
----------
-.. autosummary::
-   :toctree: generated/
-
    pipeline.make_pipeline
 
 .. _metrics_ref:
@@ -130,8 +135,6 @@ Metrics
 
 .. currentmodule:: imblearn
 
-Functions
----------
 .. autosummary::
    :toctree: generated/
 
@@ -152,8 +155,6 @@ Datasets
 
 .. currentmodule:: imblearn
 
-Functions
----------
 .. autosummary::
    :toctree: generated/
 
@@ -169,8 +170,6 @@ Utilities
 
 .. currentmodule:: imblearn
 
-Functions
----------
 .. autosummary::
    :toctree: generated/
 

doc/whats_new.rst

@@ -42,6 +42,9 @@ API changes summary
   errors. By `Guillaume Lemaitre`_.
 - creation of a module `utils.validation` to make checking of
   recurrent patterns. By `Guillaume Lemaitre`_.
+- move the under-sampling methods in `prototype_selection` and
+  `prototype_generation` submodule to make a clearer dinstinction. By
+  `Guillaume Lemaitre`_.
 
 
 .. _changes_0_2:

imblearn/under_sampling/__init__.py

@@ -3,17 +3,18 @@
 a dataset.
 """
 
-from .random_under_sampler import RandomUnderSampler
-from .tomek_links import TomekLinks
-from .cluster_centroids import ClusterCentroids
-from .nearmiss import NearMiss
-from .condensed_nearest_neighbour import CondensedNearestNeighbour
-from .one_sided_selection import OneSidedSelection
-from .neighbourhood_cleaning_rule import NeighbourhoodCleaningRule
-from .edited_nearest_neighbours import EditedNearestNeighbours
-from .edited_nearest_neighbours import RepeatedEditedNearestNeighbours
-from .edited_nearest_neighbours import AllKNN
-from .instance_hardness_threshold import InstanceHardnessThreshold
+from .prototype_generation import ClusterCentroids
+
+from .prototype_selection import RandomUnderSampler
+from .prototype_selection import TomekLinks
+from .prototype_selection import NearMiss
+from .prototype_selection import CondensedNearestNeighbour
+from .prototype_selection import OneSidedSelection
+from .prototype_selection import NeighbourhoodCleaningRule
+from .prototype_selection import EditedNearestNeighbours
+from .prototype_selection import RepeatedEditedNearestNeighbours
+from .prototype_selection import AllKNN
+from .prototype_selection import InstanceHardnessThreshold
 
 __all__ = [
     'RandomUnderSampler', 'TomekLinks', 'ClusterCentroids', 'NearMiss',

imblearn/under_sampling/prototype_generation/__init__.py

@@ -0,0 +1,10 @@
+"""
+The :mod:`imblearn.under_sampling.prototype_generation` submodule contains
+methods that generate new samples in order to balance the dataset.
+"""
+
+from .cluster_centroids import ClusterCentroids
+
+__all__ = [
+    'ClusterCentroids'
+]

imblearn/under_sampling/cluster_centroids.py renamed to imblearn/under_sampling/prototype_generation/cluster_centroids.py

@@ -13,7 +13,7 @@
 import numpy as np
 from sklearn.cluster import KMeans
 
-from ..base import BaseMulticlassSampler
+from ...base import BaseMulticlassSampler
 
 
 class ClusterCentroids(BaseMulticlassSampler):

imblearn/under_sampling/tests/__init__.py renamed to imblearn/under_sampling/prototype_generation/tests/__init__.py

@@ -0,0 +1,22 @@
+"""
+The :mod:`imblearn.under_sampling.prototype_selection` submodule contains
+methods that select samples in order to balance the dataset.
+"""
+
+from .random_under_sampler import RandomUnderSampler
+from .tomek_links import TomekLinks
+from .nearmiss import NearMiss
+from .condensed_nearest_neighbour import CondensedNearestNeighbour
+from .one_sided_selection import OneSidedSelection
+from .neighbourhood_cleaning_rule import NeighbourhoodCleaningRule
+from .edited_nearest_neighbours import EditedNearestNeighbours
+from .edited_nearest_neighbours import RepeatedEditedNearestNeighbours
+from .edited_nearest_neighbours import AllKNN
+from .instance_hardness_threshold import InstanceHardnessThreshold
+
+__all__ = [
+    'RandomUnderSampler', 'TomekLinks', 'NearMiss',
+    'CondensedNearestNeighbour', 'OneSidedSelection',
+    'NeighbourhoodCleaningRule', 'EditedNearestNeighbours',
+    'RepeatedEditedNearestNeighbours', 'AllKNN', 'InstanceHardnessThreshold'
+]

imblearn/under_sampling/tests/test_cluster_centroids.py renamed to imblearn/under_sampling/prototype_generation/tests/test_cluster_centroids.py

@@ -13,7 +13,7 @@
 from sklearn.neighbors import KNeighborsClassifier
 from sklearn.utils import check_random_state
 
-from ..base import BaseMulticlassSampler
+from ...base import BaseMulticlassSampler
 
 
 class CondensedNearestNeighbour(BaseMulticlassSampler):

imblearn/under_sampling/prototype_selection/__init__.py

@@ -14,8 +14,8 @@
 import numpy as np
 from scipy.stats import mode
 
-from ..base import BaseMulticlassSampler
-from ..utils import check_neighbors_object
+from ...base import BaseMulticlassSampler
+from ...utils import check_neighbors_object
 
 SEL_KIND = ('all', 'mode')
 

imblearn/under_sampling/condensed_nearest_neighbour.py renamed to imblearn/under_sampling/prototype_selection/condensed_nearest_neighbour.py

@@ -17,7 +17,7 @@
 from sklearn.ensemble import RandomForestClassifier
 from sklearn.externals.six import string_types
 
-from ..base import BaseBinarySampler
+from ...base import BaseBinarySampler
 
 
 def _get_cv_splits(X, y, cv, random_state):

imblearn/under_sampling/edited_nearest_neighbours.py renamed to imblearn/under_sampling/prototype_selection/edited_nearest_neighbours.py

@@ -11,8 +11,8 @@
 
 import numpy as np
 
-from ..base import BaseMulticlassSampler
-from ..utils import check_neighbors_object
+from ...base import BaseMulticlassSampler
+from ...utils import check_neighbors_object
 
 
 class NearMiss(BaseMulticlassSampler):

imblearn/under_sampling/instance_hardness_threshold.py renamed to imblearn/under_sampling/prototype_selection/instance_hardness_threshold.py

@@ -10,8 +10,8 @@
 
 import numpy as np
 
-from ..base import BaseMulticlassSampler
-from ..utils import check_neighbors_object
+from ...base import BaseMulticlassSampler
+from ...utils import check_neighbors_object
 
 
 class NeighbourhoodCleaningRule(BaseMulticlassSampler):

imblearn/under_sampling/nearmiss.py renamed to imblearn/under_sampling/prototype_selection/nearmiss.py

@@ -12,7 +12,7 @@
 from sklearn.neighbors import KNeighborsClassifier, NearestNeighbors
 from sklearn.utils import check_random_state
 
-from ..base import BaseBinarySampler
+from ...base import BaseBinarySampler
 from .tomek_links import TomekLinks
 
 

imblearn/under_sampling/neighbourhood_cleaning_rule.py renamed to imblearn/under_sampling/prototype_selection/neighbourhood_cleaning_rule.py

@@ -11,7 +11,7 @@
 import numpy as np
 from sklearn.utils import check_random_state
 
-from ..base import BaseMulticlassSampler
+from ...base import BaseMulticlassSampler
 
 
 class RandomUnderSampler(BaseMulticlassSampler):

imblearn/under_sampling/one_sided_selection.py renamed to imblearn/under_sampling/prototype_selection/one_sided_selection.py

@@ -12,7 +12,7 @@
 import numpy as np
 from sklearn.neighbors import NearestNeighbors
 
-from ..base import BaseBinarySampler
+from ...base import BaseBinarySampler
 
 
 class TomekLinks(BaseBinarySampler):